Credit Card Debt and Consumption Evidence from Household-Level Data.pdf

8/17/2019 Credit Card Debt and Consumption Evidence from Household-Level Data.pdf

1/19

1

Credit Card Debt and Consumption: Evidence from Household-Level Data

By

Tufan Ekici & Lucia Dunn*

____________________* The authors are respectively: Instructor, Middle Eastern Technical University, North Cyprus andProfessor of Economics, Ohio State University, Columbus, Ohio. Corresponding author: Lucia Dunn,email: [email protected]; phone: 614-292-8071. The authors thank Serkan Ozbeklik and Stephen Cosslettfor useful suggestions.


2/19

2

Abstract

This research investigates the relationship between credit card debt and consumption usinghousehold level data. This is a departure from previous studies which have used aggregate

measures of consumption and general debt such as the Debt Service Ratio or total revolvingcredit. We use a detailed monthly survey of credit card use to impute credit card debt torespondents from the Consumer Expenditure Survey sample. In contrast to some earlier studiesusing aggregate data, we find a negative relationship between debt and consumption growth.Our work shows that a one-thousand dollar increase in credit card debt results in a decrease inquarterly consumption growth of almost two percent. Investigations are also made into effects ofdebt within different age categories and into the impact of expected income growth on the debt-consumption relationship.

JEL Code: D12, E21

Keywords: Consumption, Credit Card Debt, Household Data


3/19

3

Credit Card Debt and Consumption: Evidence from Household-Level Data

1. Introduction

Credit cards have become a major instrument for carrying out and financing purchases in

the U.S., with the average credit card debt for balance-carrying households reported to be more

than $9,0001. Furthermore, credit card debt has risen faster than household disposable income,

raising concern among policy-makers. Increased borrowing on credit cards to finance

consumption is usually seen as a stimulating factor for the economy. However, there is concern

that high levels of debt may curtail spending in the future and hence ultimately slow economic

growth. There have been conflicting findings on this issue from studies based on aggregate data.

Here we use a new approach to shed further light on the debt-consumption relationship.

We depart from previous research in this area in the following two ways. First we use

individual, monthly household-level data drawn from a new survey of credit card use in

conjunction with the Bureau of Labor Statistics Consumer Expenditure Survey, a monthly survey

focused on household-level spending. It has been suggested in the literature that aggregate

measures may obscure the impact of debt burdens and financial constraints (McCarthy, 1997). It

may thus be the case that the debt-consumption relationship can best be determined from data on

individual households over time.

Secondly, our study focuses specifically on the effect of credit card debt rather than

overall household debt. According to the 2004 The Survey of Consumer Finances,

approximately 75 percent of all households own at least one credit card, and 58% of those

holding a credit card carry a balance. The Federal Reserve Board puts current U.S. revolving

debt at more than $800 billion – mostly credit card balances. This represents a 600 percent

1 The average credit card balance among revolvers was $5,100 according to the 2004 Survey of Consumer Finances and $9,205 according to Bankrate.com.


4/19

4

increase in such debt in the last two decades. Furthermore, the American Banker’s Association

reports that the use of credit cards for small everyday purchases is growing, with U.S. consumers

currently making more than 350 billion transactions of less than $5, representing $1.32 trillion.

Since credit cards have become a major vehicle for carrying out and financing consumption, it is

important to specifically gauge the impact of this financial instrument.

To the best of our knowledge no single data set has the monthly information on both

household consumption and credit card debt necessary to address the issues of concern here.

Monthly information on spending habits of individual households is available from the

Consumer Expenditure Survey (CEX). The only publicly available data set with detailed

information on credit card debt on a monthly basis comes from Ohio Economic Survey (OES ).

Since both of these surveys were conducted on samples whose characteristics are representative

of the U.S. population, we use an imputation procedure to match their information.2

Our results show that credit card debt has a significant and negative impact on total

household consumption growth. A one-thousand dollar increase in credit card debt leads to a

decrease of almost two percent in total household consumption growth. When consumption is

separated into durable and non-durable categories, the effect is still negative but not as

significant. These findings suggest a need for greater attention to credit card debt levels and

their general economic impact.

2. Background and Previous Research

Research on credit card market in economics literature has been relatively new. One of

the earliest issues addressed by researchers has been the sticky interest rates in this market and

effects of search on interest rates (See Ausubel, 1991; Mester, 1994; Stavins, 1996; Park 1997;

2 See Dunn, et. al (2006) for sample characteristics. The Survey of Consumer Finances (SCF) has been widely usedin this literature, but the SCF appears only once in three years and is thus not suitable for this research.


5/19

5

Brito and Hartley, 1995; Calem and Mester, 1995; Crook, 2002; Kim, Dunn, and Mumy, 2005;

Kerr and Dunn, 2007). Researchers have also estimated interest rate elasticity of credit card debt

(See Gross and Souleles, 2002b; Shen and Giles, 2006). Another strain of the literature has

looked at the effect of credit card debt on default and bankruptcy (See Laderman, 1996; Ausubel,

1997; Dunn and Kim, 1999; Domowitz and Sartain, 1999; Stavins, 2000; See Gross and

Souleles, 2002a; Fay, Hurst, and White, 2002; Lehnert and Maki 2002; Agarwal, Lui, and

Mielnicki, 2003.) In this paper we focus on the effects of credit card debt on future consumption

patterns of households.

There is some existing empirical evidence that suggests a link between general household

debt and household consumption, even though basic theoretical models (e.g., the life-cycle

model and permanent income hypothesis) predict that contemporaneous variables such as

consumer credit/debt should not play a role in consumption growth. Most previous studies have

used data series from the Federal Reserve Board for their measures of aggregate debt levels –

usually the Debt Service Ratio (DSR) or total revolving credit.3 There has been inconsistency in

these findings regarding the sign of the impact of debt on consumption. Carroll and Dunn (1997)

and McCarthy (1997) have studied the relationship between household debt growth and durable

consumption and also find a positive and significant relationship between these variables.

Ludvigson (1999) finds that changes in total installment credit and revolving credit are

significantly related to non-durable and service consumption growth. Maki (2000) finds that

changes in consumer credit and delinquencies, but not the DSR, are positively related to

consumption growth. Using a new index of credit card debt, Dunn et. al. (2006) are able to show

3 The Federal Reserve has recently undertaken revisions to the DSR to make the process used in its calculation morein line with recent changes in financial markets and consumer behavior, and a Financial Obligations Ratio has beendeveloped. (Dynan et. al. 2003, Johnson, 2005).


6/19

6

that 10-14 percent of aggregate consumer durable spending can be explained by this type of debt.

The direction of the impact in this work is positive. On the other hand, Murphy (2000), using

different control variables, finds a negative relationship between lagged values of the DSR and

spending on durable goods and services. In addition, Olney (1999) finds a negative relationship

between consumption and debt during the period of the 1930s in the U.S. Finally, Zhang,

Bessler, Leatham (2006) find no significant relationship between the U.S. household debt service

ratio and GDP for the periods of 1980-2003.

Clearly, the effect of debt repayment with interest on consumption in future periods

depends on complicated assumptions about future income growth and consumption smoothing.

The existence of non-secured lines of credit with flexible repayment, i.e., credit cards, adds

another layer of complexity to the debt-consumption picture. Most measures of debt used thus

far are highly aggregate and may not be capturing subtle aspects of the debt burden of individual

households. The present paper investigates the impact of the growth of consumer credit card

debt on the growth of consumption using individual household data for both consumption and

debt. It focuses specifically on credit card debt, as credit cards have become a major method for

financing consumption in the U.S. and are more widely utilized than other debt instruments.

3. Data

To the best of our knowledge, no single data set has detailed information on both

households’ spending habits and credit card use from the same respondents. Therefore we will

match data from two surveys where the characteristics of respondents are both very close to the

U.S. national averages – the Consumer Expenditure Survey (CEX ), which has information on

spending habits of U.S. households, and the Ohio Economic Survey (OES ), which has detailed

information on credit card usage of households. Here our credit card debt is estimated from


7/19

7

OES sample, and the estimated coefficients are then used to impute the debt of the households

with the same characteristics in the CEX sample. The matching can be done finely since the

characteristics of both samples are very close to U.S. national averages.4 (See Appendix A for

details on these surveys.)

4. Econometric Methodology

We estimate the reduced form equation by including lagged values of debt as explanatory

variables for consumption growth. The equation to be estimated is as follows:

, 1 0 1 , 1 2 , , 1ln i t t i t i t i t d C b time b W b D η + + +′ ′ ′= + + + (1)

where D is the lagged debt, time includes a full set of month dummies5 . W includes the age of

the household head and its square, the change in the number of adults, and the change in the

number of children, as is standard in estimating this type of equation.6 We will impute the credit

card debt of the households in the CEX sample using coefficients from an estimation of debt of

the OES sample members with similar characteristics. We then estimate eq. (1) using the

imputed debt variable. The imputation process is explained in the next section.

4.1 The Imputation of Credit Card Debt

We estimate credit card borrowing for those households with at least one credit card in

the OES sample with a Tobit model, since credit card debt is left-censored at zero. The Tobit

model, which is estimated by maximum likelihood, has the following form:

* 2, ~ [0, ]i i i i y x N β ε ε σ ′= + (2)

4 A detailed comparison is available from the authors upon request.5 The omitted month is April 2002.6 It is conventional to use the variables in W in order to control for changes in household preferences. See Zeldes(1989), Dynan (1993), Lusardi (1996), and Souleles (2004).


8/19

8

where if * 0i y ≤ then yi=0, and otherwise yi=yi*. We thus obtain ˆ ˆ[ , ] β σ . The coefficients and

standard errors from the estimation of eq. (2) are given in Table 1. The results there conform to

the results of previous work in the literature.

Table 1: Tobit Estimation of the Credit Card Debt of Those Holding at Least One

Card in OES Sample

Variable Definition Coefficient

Standard Error

Number Children Number of children age


9/19

9

Step 1. We first estimate the probability of having a credit card in the OES sample with a

Probit model by using the same explanatory variables as in Table 1. The variable definitions are

the same as the ones used in Table 17.

Step 2. Taking the estimated coefficients from Step 1, we impute the probability of

having a credit card for the CEX sample members with the same socioeconomic characteristics,

using time dummies to align the periods of the observations.8

Step 3. If the imputed probability of credit card ownership is greater than 0.5, we assign

the value of one (havecard=1) to that respondent to indicate that he/she has a credit card;

otherwise (havecard=0)

9

. The percentage of credit card holders in the OES sample is 80%, and

the imputed percentage in the CEX is close at 78%.

After obtaining ˆ ˆ[ , ] β σ from eq. (2) and identifying card ownership in the above three

steps, we can impute the credit card debt for those holding credit cards in the CEX sample as

follows:

Predicted Debt of Cardholders in CEX =

ˆ ˆ ˆ( / )ˆˆ[ | ] ( )[ ]ˆˆ ˆ( / )

x x

E y x x x

β φ β σ

β σ σ β σ

′ ′′= Φ +

′Φ (3)

Where ,φ Φ are the standard normal probability distribution function and cumulative distribution

function respectively.

It is important to be sure that the variables in the vector x are available in both samples.

The variables in this vector are the ones shown in Table 1. We also include time dummies in the

xi vector in order to control for seasonal trends in credit card borrowing. To reduce

computational complexity, we will have twelve month dummies and five dummies for year. In

order to get the timing correct, the imputed debt corresponds to the first month of the interview

7 The results in this step conform to the previous literature. The results are available from the authors upon request.8 The CEX has multiple observations on the same households, but for Step 2 we utilize only the first observation.9 In preliminary investigations, the value 0.5 gave us the highest rate of correctly identifying the two groups.


10/19

10

quarter. For example, if we are trying to estimate the consumption growth between the first and

second quarters of 1998, i.e. January1998–March1998 and April1998–June1998, then the lagged

debt variable corresponds to the credit card debt of the household in January 1998. This ensures

that the respondent has full information about his/her debt level before making consumption

decisions in the next two quarters. Thus this also provides a test of the Permanent Income

Hypothesis where current information should have no effect on consumption decisions of the

future.

After imputing the credit card debt for the respondents in the CEX sample, we estimate

the following equation for consumption for the credit cardholders:

, 1 0 1 , 1 2 , , 1ˆln i t t i t i t i t d C b time b W b D η + + +′ ′= + + + (1’)

where ˆ D is the imputed debt. We estimate equation (1’) for four different measures of spending:

durable spending, non-durable spending, total spending, and outlay expenditures. The

definitions of these variables in the context of the CEX and an explanation of their components

are described in Appendix B. The standard errors have been adjusted to take account of the fact

that one of the regressors in equation (1’) is generated from another sample, which leads to

consistent but inefficient estimates.10 The OLS results of the estimation are presented on Table 2

10 Details of this adjustment process are available from the authors upon request.


11/19


12/19

12

Table 4 below. We see that lagged debt tends to be significant for consumption in the middle

group of age categories but not for the very youngest and oldest age groups.

Table 3. Estimated Coefficients o f the Effect of Lagged Credit Card Debt onConsumption Growth by Age Group (Cardholders Only)

Coef ficient Standard Error t -value N


13/19


14/19

14

6. Summary and Conclusions

This research has estimated the effect of lagged credit card debt on consumption growth

at the level of the U.S. household. Since no single data set contains the appropriate monthly

information on both consumption and credit card debt, we have combined two data sets drawn

from samples with similar characteristics to estimate this relationship. Our results show that

lagged debt is negatively related to total household consumption growth.

With the exception of the work of Olney (1999) and Murphy (2000), most previous

studies using aggregate data have found a positive relationship between lagged debt and

aggregate consumption growth. Since the impact of debt burdens and financial constraints may

not be adequately detected from aggregate measures, the use of household-level data may

provide a better test of the true the debt-consumption relationship. In addition, earlier studies

used aggregate consumption data from consumers economy-wide with debt information only

from those consumers who actually held debt. Thus the relationship that emerged from those

studies cannot be applied to all consumers. By using household-level data on debt and

consumption, we can control for the presence of debt as well as other relevant socioeconomic

characteristics that impact the relationship between the two. Finally, our work is unique in that it

uses specifically credit card debt, which has become a major financing instrument for household

consumption in the last two decades.

We have also explored possible individual factors that may be connected with our finding

of a negative debt-consumption relationship and find that the relationship is still negative,

although less significant in the older and younger age ranges. We have presented statistics which

suggest that this may be due to differences in credit limits and other credit card-related factors

for different-aged consumers. We have also controlled for differences in income expectations


15/19

15

and still find that the basic debt-consumption relationship is negative. These findings further

emphasize the importance of examining alternative specifications for the debt-consumption link.

It is possible that over a longer period of time, the debt-consumption relationship for

households might look different. On the one hand, with growing income and access to

alternative forms of credit such as home equity loans, consumers may be able to manage their

credit card debt so that it does not negatively impact consumption. On the other hand, the

increasing trend of personal bankruptcies suggests that credit card debt may continue to be a

problem for household consumption in the long-run. More extensive panel data sets are needed

to better assess these long-run possibilities for policy actions.

Appendix A: The Two Surveys

The CEX is a national survey of the Bureau of Labor Statistics which collects data on the

spending habits of U.S. households. We utilize the Interview Survey component in which

respondents are interviewed five times, three months apart. Different households are interviewed

starting in different months. Survey participants are asked to record the dollar amounts for goods

and services purchased during the 3-month period prior to the interview date, and the resulting

data covers about 95 percent of household expenditures.13 The final estimation includes about

85,000 observations.14 The CEX survey does not identify credit card ownership except for those

with positive debt, and this is only recorded for the first of the second and fifth interview months.

Also, the CEX does not take account of any debt repayment which may occur in the month. The

OES , however, records exact credit card debt, including zero debt, on a monthly basis. Its data

come from a survey taken each month of at least 500 households for the period January 1997

through April 2002, giving a total sample size for the OES is 40,320.

13 For more information on the description of the CEX survey, see CES Anthology (2003).14 Sample selection was done in according with the conventions set in the previous literature that has used CEX data.


16/19

16

Appendix B: Creation of the Dependent Variables

The CEX has a constructed variable with information on “total expenditures.” We use

this variable as our total consumption variable. According to the official CEX definitions, total

expenditures includes the following: spending on food, alcoholic beverages, housing, apparel,

transportation, health, entertainment, personal care, reading, education, tobacco, miscellaneous

items, cash contributions, and personal care. All of these components have subcategories. For

more information on what is included in the subcategories, please see the CEX documentation.

Finally, the spending amounts on these components for the three months prior to the interview

month are added together to create the total spending values.

Following BLS conventions, we also use the “total outlays” variable as a measure of total

consumption. The “total expenditures” variable described above included the total cost of the

items purchased in that period. Thus, if someone has financed a big ticket item and only paid a

certain portion of it within the quarter in which he/she was interviewed, the amount paid will be

included in the outlays variable. “Total spending,” however, will include the full price of the

item even though it is not yet paid in full. Thus in order to capture the actual amount of spending

for that quarter, we also use the “total outlays” variable as a measure of consumption.

Next we separate the total spending into durable and nondurable spending. Nondurable

spending includes the following: spending on food, alcoholic beverages, utilities, apparel,

tobacco, and gasoline, and motor vehicle-related expenses. Durable spending includes spending

on housing and tv/radio/sound equipment. The breakdown of total consumption into durable and

nondurable consumption is done individually by the authors and thus may vary among different

users.


17/19

17

REFERENCES

Agarwal, Sumit, Chunlin Liu, and Lawrence Mielnick. ‘Exemption Laws and Consumer Delinquency andBankruptcy Behavior: An Empirical Analysis of Credit Card Data.” The Quarterly Review of Economics and Finance 43 (January 2003): 273-289.

Ausubel, Lawrence M., “The Failure of Competition in the Credit Card Market,” The American Economic Review, vol 81, no. 1 (March, 1991), pp.50-81.

____________________ “Credit Card Defaults, Credit Card Profits and Bankruptcy.” American Bankruptcy Law Journal 71 (Spring 1997) 249-270.

Brito, Dagobert l. and Peter R. Hartley, “Consumer Rationality and Credit Cards,” Journal ofPolitical Economy, vol. 103, no. 4, (1995), pp. 400-433.

Brown, Sarah, Gaia Garino, Karl Taylor, and Stephen W. Price (2005). “Debt andFinancial Expectations: An Individual and Household Level Analysis.” Economic Inquiry, 43(1), January 2005, 100-120.

Calem, Paul S., and Loretta J. Mester, “ Consumer Behavior and Stickiness of Credit CardInterest Rates,” American Economic Review, 85(5), December 1995, 1327-1336.

Campbell, J., and Mankiw, G. (1990). “Permanent Income, Current Income and Consumption,” Journal of Business and Economic Statistics, 8, 265-279

Carroll, Christopher D. and Wendy E. Dunn. “Unemployment Expectations, Jumping (S,s)Triggers, and Household Balance Sheets,” NBER Macroeconomics Annual 1997.

Carroll, Christopher (1997). “Buffer Stock Saving and the Life Cycle/Permanent IncomeHypothesis,” Quarterly Journal of Economics, volume CXII, pp. 1-56

Crook, Jonathan N.. “The Demand for Household Debt in the USA: Evidence from the 1995Survey of Consumer Finance.” Applied Financial Economics 11 (2001): 83-91.2001.

________________ “Adverse Selection and Search in the US Bank Credit CardMarket.” University of Edinburgh Credit Research Center Working Paper No.01/1, March 2002.

Deaton, Angus (1991). “Saving and Liquidity Constraints”, Econometrica, (September, 1991)

Vol. 59, No. 5, 1221-1248

Domowitz, Ian and Robert L Sartain. “Determinants of the Consumer Bankruptcy Decision,”The Journal of Finance, February, 1999, v.54, pp. 403-420.

Dunn, Lucia F., Tufan Ekici, Paul J. Lavrakas, and Jeffery A. Stec. “The Effect of Credit CardDebt on Aggregate Consumption: Evidence from a New Debt Index” Ohio StateUniversity Department of Economics Working Paper, 2006


18/19

18

Dunn, Lucia F. and TaeHyung Kim. “An Empirical Investigation of Credit Card Default,”Working Paper # 99-15, Ohio State University. October 1999.

Dynan, Karen E. (1993). “How Prudent are Consumers?” Journal of Political Economy, 101, pp. 1104-1113.

Dynan, Karen, Kathleen Johnson, and Karen Pence. “Recent Changes to a Measure of U.S.Household Debt Service”, Federal Reserve Bulletin, October 2003, 417-26.

Ekici, Tufan. “Do the Expected Real Interest Rate and Consumer Confidence Matter for CreditCard Borrowing?” Ohio State University Working Paper , 2006

Fay, Scott, Erik Hurst, and Michelle J. White. “The Bankruptcy Decision” American Economic Review 92 (June 2002): 706-718.

Gross, David B. and Nicholas S. Souleles (a). “An Empirical Analysis of Personal Bankruptcy

and Delinquency.” The Review of Financial Studies 15 (spring 2002): 319-347.

_____________________________________(b). “Do Liquidity Constraints and Interest RatesMatter for Consumer Behavior? Evidence from Credit Card Data.” Quarterly Journal of Economics 117 (2002): 149-85.

Jiang, Saihong. “Credit Card Debt and Payoff: A Lifecycle Analysis.” Ohio State University Mimeo , 2006

Johnson, Kathleen W., “Recent Developments in the Credit Card Market and the FinancialObligations Ratio,” Federal Reserve Bulletin (August, 2005), pp. 473-486.

Kennickell, Arthur B., and Martha Starr-McCluer, "Changes in Family Finances From 1989 to1992," Federal Reserve Bulletin, (October, 1994), v. 80, pp. 861-882.

Kerr, Sougata, and Lucia Dunn (2002). “Consumer Search in the Changing Credit CardMarket.” Ohio State University Department of Economics Working Paper No.02-03, September 2002.

Kim, TaeHyung, Lucia F. Dunn, and Gene E. Mumy. “Bank Price Competition and AsymmetricConsumer Responses to Credit Card Interest Rates.” Economic Inquiry 43 (April 2005):344-353.

Laderman, Elizabeth. “What’s Behind Problem Credit Card Loans?” Economic Letter 96,Federal Reserve Bank of San Francisco (July 1996).


19/19

19

Lehnert, Andreas and Dean M.Maki. “Consumption, Debt and Portfolio Choice: Testing theEffect of Bankruptcy Law.” Finance and Economics Discussion Series 2002-14, TheFederal Reserve Board (2002).

Ludvigson, Sydney. “Consumption and Credit: A Model of Time-Varying LiquidityConstraints,” The Review of Economics and Statistics, August 1999, 81 (3): 434-447.

Lusardi, Annamaria (1996). “Permanent Income, Current Income and Consumption: Evidencefrom Two Panel Data Sets,” Journal of Business and Economic Statistics 14: 81-90.

Maki, Dean M. “The Growth of Consumer Credit and the Household Debt Service Burden,” Board of Governors of the Federal Reserve System, Finance and Economics Discussion

Paper : 2000/12.

McCarthy, Jonathan. “Debt, Delinquencies, and Consumer Spending”, CIEF, Federal ReserveBank of New York, February 1997, Volume 3, Number 3

Mester, Loretta J. (1994). “Why are Credit Card Rates Sticky?,” Economic Theory, 4,505-530

Murphy, Robert G. “Household Debt and Consumer Spending”, Business Economics, Jul 98,Vol. 33 Issue 3, p38.

Olney, Martha L. (1997). “Avoiding Default: The Role of Credit in the ConsumptionCollapse of 1930,” The Quarterly Journal of Economics, February 1997, pp. 319-335.

Park, Sangkyun, “Option Value of Credit Lines as an Explanation of High Credit Card Rates.”FRB of New York Research Paper 9702 (February 1997)

Shen, Kaili and David E. Giles (2006). “Rational Exuberance at the mall: Addiction to Carryinga Credit Card Balance,” Applied Economics, 2006, 38, 587-592.

Souleles, Nicholas S. (2004). “Expectations, Heterogeneous Forecast Errors, and Consumption:Micro Evidence from the Michigan Consumer Sentiment Surveys,” Journal of Money,Credit, and Banking, February 2004, v. 36, iss. 1, pp 39-72

Stavins, Joanna. “Can Demand Elasticities Explain Sticky Credit Card Rates?” New England Economic Review (July/August 1996): 43-54.

______________. “Credit Card Borrowing, Delinquency, and Personal Bankruptcy.” New England Economic Review (July/August 2000)

Zhang, Yin and Guang Hua Wan (2004). “Liquidity Constaint, Uncertainty and HouseholdConsumption in China,” Applied Economics, 2004, 36, 2221-2229.

Zhang, Jin, David A. Bessler, and David J. Leatham (2006). “Does Consumer Debt CauseEconomic Recession? Evidence Using Directed Acyclic Graphs,” Applied EconomicsLetters, 2006, 13, 401-407.

Credit Card Debt and Consumption Evidence from Household-Level Data.pdf

Documents

Transcript of Credit Card Debt and Consumption Evidence from Household-Level Data.pdf